XX Geeky hesitates and continues

[crisnee]

Hi Peter,

The following has happened several times. My current signature is valid for this issue. The system had been newly rebooted.

The music hesitates for a second or two and then continues in perfect form, sounding very good. The most recent time it happened I was listening to four tracks, total time just short of an hour. XX hesitated once during the first track about 5 - 8 minutes in, and again during the last track, maybe 10 minutes before the end. I detected no other problems, no ticks etc.

It was running unattended so I can't give you anymore details. Just thought you'd like to know.

Chris

[PeterSt]

Very well put (question) ...

Memory is always "virtual"; At this first stall the next part of the track will be read into memory. Careful : this depends on the length of the track, so only tracks of 10 minutes or so (taking your 2x AP in account) will exhibit this theoretically;
Although for your current MC it can't happen really, let's say that at that time it needs "allocation" of the memory the first time and it will take more time.

Towards the end again a special situation occurs (mind you, (you) Unattended only), because there's a decision point : shall we read the usual SFS, or shall we read a little more because we'd be done with it (otherwise leaving that little extra for a last time). Here too, similar will happen as in the beginning (suddenly more real memory is needed).

With Mixed this for sure has impact, because there the "allocation" has to be done right from the (native or so) base. It will take much more time. Now :

Since I can use Mixed 500 without anomalies, your MC not working 100% should show a slowish system. No no, don't talk about 3GHz etc. ... it just will. But where and how ? don't ask me.
You may have read this, because it pretty much covers for general guidelines : A Guide to Glitchless Playback. But if you didn't, please do.

But watch out :
It wouldn't be the first time that someone like you would report : after a reboot all is well ! Yea, even if you just started out with a reboot. *That* part I don't know, and all I can tell is that it can happen to me too (but actually a longer time ago, not with 0.9z-4-1 so far).
Anyway, read that link if you didn't before.

Peter

[crisnee]

I'm just bringing this up again in case others are having similar problems. Check out the first post for more details.

Peter, I continue to have the same problem sporadically, no clearly apparent pattern (hesitation, sometimes a split second, sometimes a couple of seconds and then the music continues faultlessly). Some sessions it doesn't happen at all. Length of file doesn't seem to be the issue, nor if it's the first in the session. Yesterday I listened to a Brahms symphony (longish files) plus several 30 minute files, no problems. I listened to short files, no problem. And you guessed it, at other times had problems with both, sometimes several during a session. Then with no reboots after said problems, no problems.

Over the course of these sessions I've reduced the SFS in Mixed Continuous from 350 to 300 to 240. Changed to no upsampling just to experiment, no change. Seemingly random hesitations continue.

All this is no big deal to me, as they don't happen often and once done the music continues, but still... If it matters to you to know what's going on, give me a simple plan of attack, if not, or if you think I have an isolated problem, that's fine too.

Chris

[PeterSt]

Ok Chris, this is not easy. All we can work with is the fact that someone like me doesn't have any of these problems, BUT they sure can be there. And, when such a someone like me solves it, he will not be oriented towards others who may have the same problem, because you won't think about that in the first place, and furthermore all problems can be seen as unique. And maybe they even are really.

Two side notes, but maybe they are useful :

When I used W7 RTM (like you do according your sig), I only had problems and problems to let all run as expected, which plainly worked for Vista before, and in the end was tackled for W7 just the same. But, by means we should use for SQ, and now instead were used to let all play fluently. Maybe this was before your time, but hundreds of posts have been spent on that phenomenon alone (almost everybody had those problems when going from Vista to W7 (well over a year ago I mean).

When I again went back to W7, now SP1, things behaved different, to the sense of W7 SP1 plainly stalling the sound on behalf of ... ? catching up or something. It seems to trade distorted sound for no sound at all when it actually would be distorted. Something like that.
What I can not tell though, whether this would have happened to W7 RTM just the same, because some months of development has been in between going *back* to Vista, and when everybody did, going to W7 SP1 because it was there. All I can say is that the W7 SP1 behaviour from now, is very different from Vista.

Funnily enough, after reading your first post about this, I thought of referring to this : http://www.phasure.com/index.php?topic=1635.0;all
which I did not because it has been an offline operation, and I thought it could be inappropriate. I thought of asking Bert to jump in, and tell you about what stupid actually unrecognizeable things can happen, which get solved by reinstalling the OS and some more things. Since now that topic coincidentally emerged, I can just as well stick to that original idea, so  maybe Bert can tell about it in a few lines.
Anyway what this is about is the (as it looks) very same problem, which for me is nothing less or more than "a" user with "some" system, and it doesn't behave - but I don't know what happened to that system. So what to do ? No guidelines here, except for saying something like "reinstall the OS and hope it's okay after that'. Btw, that this is only the start of things, expresses beautifully in that topic I referred to, because some things may be solved, byt not all. Or not right away; XX Settings are used the same as before, while they actually now should be different, and it's all a bit of being in the blind for a user. Why ? well, because still that disk light is flashing all the time, and it should not. It may be totally harmless, but everybody can imagine that even the slightest load of whatever it is which can be avoided, WILL limit your possibilities to some extend, and where I say I can play glitchless with an SFS of 240, you can not reach that because of such a thing.

The real problem might be :
Now we manged to be able to play with a way more higher SFS than before, BUT, this is only a stupid physical thing - and we must realize that it now needs more things to be optimally working than before. Ehm, *if* it is about this at all, and so it is just an example.

Anyway for now, and sorry I don't have anything better (but please read that other topic too) : at this moment all I can say it is just possible because I can do it too. This with a most probably way higher load than you can achieve, which in your case can be about 24/88.2 (HiFace, 2x) and in my case is 32/384. I can't tell by heart what your Q1=3 implies, but I guess it will be something like 128 which is the same as mine. Thus, 5.8 times more load in my case. Processor speed is about the same.

Try the HFace at 2048 or 4096 ?

Peter

[crisnee]

Funnily enough, after reading your first post about this, I thought of referring to this : http://www.phasure.com/index.php?topic=1635.0;all
which I did not because it has been an offline operation, and I thought it could be inappropriate.

Peter, It is exactly because of that post that I posted more on this topic. It's hard for me to believe that no one else is having a similar problem. Since it's not much of a problem people might skip mentioning it. So I thought if I post again someone else with the problem might chime in. Well I was wrong, either no problem or no chiming.

Anyway as to Win7 Sp1. I hadn't updated to Sp1 because my Win7 had most of Sp1 in it already. I had just re-installed it very recently and then updated it to the beginning/middle of Feb. I then took the pc off line.

I pretty much don't and haven't run anything on it but the very basics and XX so it's pretty much pristine. Being a good camper I decided to put the pc on-line temporarily and install sp1 (of the approximately 80mg-700mg size range of sp1 my Win7 only needed the minimum).

So, after all that, The same or very similar problem persists, go figure. And I'm the only one?

I've come across another very minor and file specific problem. It won't play a particular flac file, (some kind of problem in header)  and if I convert said file to a .wav it won't play it either and lists the time incorrectly (correct time 1.29, lists it as 9 minutes something). Needless to say since I am able to convert it, (it plays in other software too) the file is not corrupt. Any ideas?

Chris

[Calibrator]

I had my first occurence, that I can recall, of a very brief pause last night ( maybe 1/4 second if that ), while listening to ( I think ) track 4 of Grover Washington Jr's "Winelight" in 24/192 format. I did have my SFS at 500, which I have had set like that for a week or more now, but I've dropped it back to 480 to be a little more conservative. Not sure whether that might have been the cause but I'll continue to monitor.

Rest of parameters as per signature below.

Cheers,

Russ

[PeterSt]

I don't even know whether I told it, but I have it too, sometimes. How can I otherwise know about it (see earlier post).
The example was yesterday, and maybe Russ and I can find out whether it was at the same time. Star radiation.

But seriously, I do have it some times, but it must be less then once a week. Yesterday was such an occasion (sound stayed away for 5 seconds). But keep in mind (and I don't think you read that - or I didn't say it explicitly in *this* topic) : this is W7 SP1.

When such a thing happens once per week - or even once par day, personally I wouldn't complain. But, when I was up to the best means of playback, I would mention it. Well, I did (I guess it was in the first post of the W7 SP1 topic). But many (MANY) people don't feel obliged to do so (and nobody is of course), and even helping out may not trigger them. In the mean time they have the problem themselves and it won't be solved (say, because of themselves).

All 'n all, to me it doesn't say much that nobody responds (except for Russ who always does that), and to you (too) it should not tell that you are alone. Still the problem is out of my control because for 100% sure it will be your "PC". Guidelines to solve it I don't have, and for example, after many emails spent with in this occasion Bert, he solved it, but doesn't know (or doesn't tell me) how.

"His master's voice" may do a lot, and chances are fair that if I'd look to your system I can tell right away which things are "wrong" or inconsistent. It's just general PC stuff, and honestly I don't know of anyone who understands even a little about it. I'm not pointing at you because I just don't know, but chances will be rather small. We, where I am typing righyt now, as an ERP company were ahead of everything and all back in 1987 were it for what PC's can do. Still, it doesn't matter which personnel's PC I sit down at, it's *always* so much more without response than what I would allow myself to work with, that I just can't understand that even people with the explicit skill, don't get what's possible with PC's. Or what's wrong with them. They all work 20 years here. Don't let them read this.

But :

What seems logical to me, is that I could be over driving things. And let's say I for sure did when everybody would have problems like this. On the other hand, I'd say it would rain messages like yours, or like the one from Bert. Still, things are temped by going up to limits, which limits are feaseable for me, but may not for others. Maybe I'm not thinking about that enough, but what to do when my PC is allright from the start, and when it's not I do something about it. Can't prevent myself from that. But the example is the "Tweak to dead" topic, which was about the first thing I saw necessary, just don't liking flashing disk lights all the time. And next : what to do if many of the others I can't see from here, don't mind at all - put a sticker over it so to speak, and next have such a problem.

No, this doesn't help you one bit. But hopefully it can encourage people to speak up about even the slightest problem they encounter. Everybody should.


About your problem with that file ... it *will* be wrong, but depending on how software deals with it, it is not encountered. You could blame me to deal with it wrongly or badly, but the file just will be wrong. Something inconsistent in the header. It can theoretically be a bug on my side, but that is hard to believe; too many "test data" on my side, and all just works. When it does not, I can find a problem in 100% of cases in the file. On this matter, I will be happy to look at it to ease your mind if necessary, *or* to find something wrong which is just solveable (if I'd count the tweaks for such situations I might reach the 100).

For now, regards,
Peter


PS: Take care that all your WiFi, IR (keyboards) and other transmitting stuff is Off. And, we know by now that even Bluetooth is bad.

[PeterSt]

I am just thinking ( ) ...

I ever made some "OS analysis" tool for myself, which I think (!) could help out with what could be going on here. This code could be activated.

The point is though, this code by itself may cause the "stalls" already, because it will create quite some overhead. So at this moment I am not sure how to implement it, without "degrading" in the first place.

Peter

[crisnee]

About your problem with that file ... it *will* be wrong, but depending on how software deals with it, it is not encountered. You could blame me to deal with it wrongly or badly, but the file just will be wrong. Something inconsistent in the header. It can theoretically be a bug on my side, but that is hard to believe;

PS: Take care that all your WiFi, IR (keyboards) and other transmitting stuff is Off. And, we know by now that even Bluetooth is bad.

Peter, The file is one in a series of 30 ripped from one cd in one sequence, no errors reported, blah blah blah. I post these things not because I'm terribly concerned, but because I'm curious, why the header would be corrupt (sort of) in such a situation. It's one in a series in which all the others are fine, and all including that one are accessible by other programs. Just odd. I've worked with computers since before the pc existed, so I have an electronically instilled curiosity. And my computers (the ones that matter) are kept running smoothly and with a minimum of "stuff," installed.

As to wifi and wireless, I don't have that sort installed on the "audio" pc.

Peter, don't feel obligated to answer my "curious issues," at least not in a hurry; Ones I term "minor." Just let me know that you don't have time for a particular one, when you don't. I just feel you should know for your own and XX's sake, it might help in the long run.

As to the hesitations and "over driving the system." The odd thing is that when I reduce the load (no upsampling, lower the sfs) nothing seems to change, the occasional hesitations continue. Sometimes there are three in an album, sometimes none in several, often they're very short (split second) very occasionally once every five? days, there's one 3-7 seconds long. Happy listening, don't fall in the crevices.

Chris

[PeterSt]

Thank you Chris.

Really, send me that file please. I will have a look, and adjust things when necessary. But the least we should do is find out what's wrong, especially when you yourself can't think of anything.

About the "stalls" and in inconsistency regarding the load;
Can you try to let spin your disks for 15 minutes or so ? So, let them spin down after 15 minutes (or 10, or 20). Be careful with this if you have many and an underrated cooling system (or even PSU), like I have ...


One more remark : W7 SP1 deals differently with the I/O's to the OS disk. There are less. So, my spinning disk spins down (set at 2 min), while before (W7) it did not ever. Now, I also have the feeling that there's some lag in spinning it up. This easily can take 10 seconds, but sometimes it may even be way longer. In the end this seems to reflect everything, and even hovering the startbutton will not make that active during that period. This doesn't happen during playback (with ME), but for example when I don't login right away, it will take me 30 seconds when I do it later.

These are load-unrelated things *and* dedicated to SP1 ...

Peter

[crisnee]

About the "stalls" and in inconsistency regarding the load;
Can you try to let spin your disks for 15 minutes or so ? So, let them spin down after 15 minutes (or 10, or 20). Be careful with this if you have many and an underrated cooling system (or even PSU), like I have ...

Do you mean set the automatic disk spin down (ie disk sleep) to 20 minutes or so? I think that's what I have it set for if that's what you mean. I'll check it later when I get on that computer.

As to the problem file, I'll send it later too, as I have to utilize the sneaker net    to get it to an on-line system. 

Chris

[PeterSt]

FileMail ?

[Suteetat]

Hmmm... I think I just ran into similar problem today, may be not. Music seems to stutter for a couple of seconds here and there. No change in the setting over the past several weeks except for migrating XXHighend from Ramdisk over to C: drive (SSD).
Also the last couple of days, Norton antivirus has been trying to remove engine 3 file as it reported high risk activity or something (I have Norton in this computer for several months now and this is the first time Norton is trying to remove XXHighend) and I asked Norton to unquarantine the program. The stuttering problem started to recurred. Anyhow, I just reinstalled XXhighend and Norton does not do anything to XXHighend anymore and
stuttering problem seems to disappear for now.
 

[PeterSt]

Apart from the virus scanner which always can (actually will) create problems, don't underestimate the speed of a RAMDisk vs. any other type of "persistant" storage. So, of course, if your system is a little up to limits but all is ok while using a RAMDisk, all will go over limits at using that other type of storage.

[crisnee]

Peter, Ok, finally, my sneaker net mail arrived. I'll send the file via pm.

As to my hard drive settings. Turned out it was set to 10 minutes 'til spin down. Changed it to 20 minutes. So, later that day I played some music, about 6 or 7 Fleetwood Mac tracks, all pretty short. I played them, all good, then I played them again having changed nothing. In the middle of the 3rd to last track the music stopped. I thought "oh well here we go," except there was no go. It just stopped. After waiting about 20 seconds I went to look at the screen. There was the engine stopped, must close program message.

However, it was not stopped, at least not completely, as the time bar was still moving. So I decided not to close XX, but rather I hit stop button and then start the music again; lo and behold, the music played (from the beginning, not from the stopped place). So, that's that. I don't know if it's related to the hesitation but...

I'll report more about the 20 minute hd setting when I've played music for a few days.

FYI (should it matter), much of the time I've been playing XX in attended mode, as my computer has almost no unnecessary services (3 maybe?) running; it's used pretty much for nothing but audio. It takes about 2 seconds for XX to shut them down when I do run it unattended.

Chris